I cannot get the Standardize filter to ignore the class. I have 5
attributes, all numeric (including the class). I tried using the
-unset-class-temporarily parameter, but it didn't work. And I am selecting
the right class on the bottom right corner of the explorer gui.
Hi, I use the API of weka for build my classifier, and I need know if is
possible that I can manipulate the instances that are ocupated in the
training in cross validation??
Evaluation ev = new Evaluation(datosOriginals);
J48 tree= new J48(); // my base classifier
In this part
I want use the instances of training in each validation.
It is possible???
Recently I've been asked about the validity of using Root mean square for
evaluating probabilities (response from 0 to 1).
Despite I've found some previous posts regarding similar subjects I couldn't
find a clear answer for this.
Can someone provide further insight?
Thanks in advance,
View this message in context: http://old.nabble.com/Root-mean-square-for-evaluating-binary-response-tp287…
Sent from the WEKA mailing list archive at Nabble.com.
I am using Weka, the Explorer. I know that we can get the cluster assignment for each instances or observations by visualizing the results and save it it to a file. How about the distance? Can we get the distances measured for each observations to the center of the clusters (or the similarity measure for each observations to the rest of the cluster member) by using the Explorer? If yes, how can I do it?
Thank you so much in advance
The New Busy think 9 to 5 is a cute idea. Combine multiple calendars with Hotmail.
A lot of the common examples are explained using the J48 classifier or
the M5P classifier. Are these good for textual classification? Also I
saw that these classes dont seem to implement the UpdateableClassifier
interface, that means these can be used only for one shot training, isnt it?
I want to be able to classify tokenized text and am looking for a
classifier which can be trained on an incremental basis. Can anyone
suggest which classifier is appropriate for such a scenario?
I was wondering if somebody tried to implement convolutional neural nets
with Weka and how did you do it?
My guess would be to just use the MultiLayerPerceptron class and encode the
convolution at each layer. Any hints?